New beginnings: The right to equality and early childhood care and education

Abstract While South Africa has seen important advances in the provision of early childhood care and education (ECCE), about 3.2 million children still lack access to any programme. Problems of access and quality are most pronounced in the poorest communities. Even before Covid-19 forced many providers to close, these programmes were overcrowded, with poor infrastructure, and an under-paid and under-qualified workforce. ECCE is crucial for a child’s development, meaning that these inequalities are amplified in school and later life. This has knock-on effects for caregivers, particularly women, and their ability to access quality work. This article argues that the right to equality can be mobilised both in relation to the Constitution of the Republic of South Africa, 1996 and international law to address these disparities. By using a framework of substantive equality, we conclude that poverty, gender and race are potential grounds for discrimination both directly and indirectly. We further propose that resource-based justifications for limiting this right are unacceptable when budgets permit unequal resource distribution and contravene a government’s positive duty to fulfil the right to equality.


Introduction
While recent years have seen important steps forward in the provision of early childhood care and education (ECCE) in South Africa, nearly 3.2 million children under five still have no access to any programme. 1 Services are frequently seriously underresourced. 2 It is in the poorest communities that problems in both access and quality are most pronounced. Even before Covid-19 forced many providers to close, there with a commitment to equality, both as a value and as a right, it is worth exploring how the right to equality can be developed and applied to ECCE.
Part 2 of this article sets out the challenges in South Africa in relation to ECCE, and provides data on the deep inequalities in access. Part 3 asks how the right to equality can be mobilised both in relation to the Constitution and international law, with a particular focus on the International Covenant on Economic Social and Cultural Rights (ICESCR), the Convention on the Rights of Persons with Disabilities (CRPD) and the Convention on the Rights of the Child (CRC). Figure 1 summarises the relevant pieces of legislation covered in this article. We begin in Section 1 of Part 3 by setting a framework of substantive equality against which to assess what the right to equality should mean in this context. Sections 3.2 and 3.3 consider poverty, gender and race as potential grounds for discrimination in relation to ECCE. The fourth section explores how the concept of indirect discrimination might be developed, followed by an analysis of resource-based justifications for limiting the right to equality. In the final section of this Part, we explore positive duties to fulfil the right to equality. Here we talk about the care and education aspects of ECCE together, in the recognition that education and development have a reciprocal relationship.  of comprehensive, universally available and equitable ECCE services by 2030. 14 The ultimate goal is universal access to quality ECCE services for all children from conception until they enter formal school or turn seven. One year of pre-primary (Grade R) for five-year-olds is in the process of becoming compulsory and already has close to universal enrolment, although the goal is for two years of compulsory pre-primary. 15 There is still, however, a long way to go before these goals are achieved.

Context and challenges
A major obstacle has been that to receive public funding, ECCE centres must be registered through the Department of Social Development (DSD). To be registered, they must fulfil certain standards with regard to safety, bathroom facilities, kitchens or areas used to prepare food, indoor and outdoor space and isolation rooms for sick children, among others. 16 Requirements include at least one meal per day, meeting nutritional and sanitary standards; a staff-child ratio ranging from 1:6 for the youngest children (aged one to eighteen months) to 1:30 for older children (five to six); and staff trained in skills such as first aid. 17 Funding takes the form of subsidies: for example, R17 per child per day (as of 2020) for each child from a family with income below the means test. Additionally, centres have their early learning programmes assessed and certified by the Department of Basic Education (DBE), and with this certification come subsidies for teacher stipends and training, as well as requirements for curricula and quality. However, most community-based ECCE centres struggle to meet these requirements and have consequently not been assimilated into the system and failed to access government funding. 18 The responsibility for early childhood development (ECD) formally shifted from the DSD to the DBE in 2022. This transfer is promising, but policy and practice guidelines that improve on the current provision have yet to be finalised. It is also not clear how the DBE will work with the Department of Health and the DSD to ensure an integrated approach to early childhood development.

Pervasive inequalities
Varying in terms of formality, quality, subsidy and staff training, ECCE services range from playgroups with little focus on learning, day mothers who care for small groups in their homes, to informal and formal ECCE centres. Pre-Covid-19 data shows only half of three-to four-year-olds participate in any early learning programme, and only half of these children attend programmes of sufficient quality to improve learning, with wide disparities in enrolment and attendance across provinces. 19 The ECD Census 2021 estimates that only a third of children aged three to five are enrolled in an ECD programme. 20 ECCE programmes tend to exclude children living in poverty as well as those living in rural areas or informal settlements. For example, the latest data estimates that 62 per cent of children in urban areas are enrolled in an ECD programme, compared to 38 per cent in rural areas. 21 This is partly because services are either provided by the private sector with unaffordable fees or inadequately funded not-for-profit organisations that struggle to deliver quality services. 22 In addition, access to documentation such as birth certificates is a challenge for both families and centres. The majority of formal ECD centres require a birth certificate for registration since centres are not able to claim government subsidies for children without birth certificates. In more informal centres, particularly those close to country borders, where migrant children may be attending, up to 93 per cent of children aged zero to two have been found to have no birth certificates. 23 Although this is not strictly adhered to in some of these informal centres, the need for a birth certificate can nevertheless constitute a significant barrier to attendance and centre quality, since subsidies are a key resource.
Children living in poorer households are less likely to be exposed to high-quality early learning programmes, if any. In 2016, almost half of the children in lower household-income quintiles did not attend any educational centre. 24 These children are also potentially less likely to receive stimulation at home. The 2016 General Household Survey found that population group and household income are both limiting factors for stimulation received in the home. 25 A large percentage of children get no stimulation at home or in formal learning activities. 26 A greater proportion of girls than boys are likely to be enrolled in home-based playgroups. However, boys are more likely than girls to attend Grade R in a primary school setting, suggesting that girls may be kept home while boys are sent to school at this early age. 27 The quality of informal ECCE programmes and centres is relatively poor. ECCE facilities face scarcity of learning materials and resources, especially within the classroom setting, minimal funding, lack of qualified teachers, inadequate security for children and poor toilet amenities. 28 Many ECCE facilities function without basic infrastructure, such as running water, access to electricity or suitable sanitation. 29 The 2013 national audit of ECCE centres found that just over half of centres could not be fully registered due to infrastructure issues. 30 Although registered centres tend to fair better in all aspects compared to unregistered centres, registered centres can experience similar quality and resource challenges, and the socioeconomic location of a centre seems to be a driving factor. Centres in the lowest quintile areas are less likely to have a broad range of learning material, opportunities for play or to 21  encourage child-driven learning. 31 Practitioners and playgroup facilitators are in short supply and under-skilled. Half (48 per cent) of current ECCE practitioners are unqualified and need training. 32 One 2009 study shows that in a single province, only 35 per cent of practitioners responsible for infant and toddler classes, and only 47 per cent of practitioners responsible for older children, have any form of ECCE qualification. 33 One consequence of this low exposure to quality ECCE services is that by Grade 4 (age nine to ten), only 23 per cent of children can read for meaning. 34 South Africa spends only 1 to 2 per cent of its total education budget on early learning programmesnot enough to power better educational outcomes. 35 3. Mobilising the right to equality: South African and international human rights law SA has a rich constitutional jurisprudence on children's rights, but, until Covid-19, it had not fully engaged with ECCE. The two Covid-19-related cases 36 that came before the High Court did engage directly with ECCE, but not fully in a rights-based way. This raises the question of how to ground ECCE in the Constitution. Ally et al argue that it should be based on the right to life, opening the door to a holistic approach to ECD. 37 This article examines the further and complementary possibility of using s 9, or the right to equality, a possibility briefly raised by Julia Sloth-Nielsen but not explored in detail. 38 The right to equality in s 9 of the Constitution has not been prominent in the right to education litigation, despite the depth of inequalities remaining unaddressed after apartheid. Nevertheless, the particular need to extend and improve ECCE for the poorest and most marginalised communities was emphasised in both Covid-19 cases, especially the second, 39 which ordered subsidies to be paid to all registered centres throughout lockdown whether operational or not. Moreover, in 2018, the United Nations (UN) Committee on Economic, Social and Cultural Rights (CESCR) expressed its concern that, in spite of the South African government's efforts, participation among children in low-income families in early education remains low. It therefore recommended that South Africa guarantee high-quality early education for all children, especially for disadvantaged families. 40 This clearly sets the scene for a potentially important role for s 9. A particularly powerful endorsement of such an 31 DBE (note 20 above). 32  approach came from the Constitutional Court in Juma Musjid, in which Nkabinde J stated: 'Today, the lasting effects of the educational segregation of apartheid are discernible in the systemic problems of inadequate facilities and the discrepancy in the level of basic education for the majority of learners'. 41 This statement was made in the context of basic education, rather than ECCE. It is therefore worth exploring more concretely how s 9 might be applied to address these disparities in South Africa. The achievement of equality is a core value of the Constitution, permeating all the provisions of the Bill of Rights. 42 The right to equality itself is set out in s 9 of the Constitution. Under s 9(1), everyone is equal before the law and has the right to equal protection and benefit of the law; while ss 9(3) and 9(4) prohibit the state and all persons from unfairly discriminating, directly or indirectly, against anyone on one or more of a range of grounds, including, race, gender, sex, pregnancy, marital status, ethnic or social origin, colour, sexual orientation, age, disability, religion, conscience, belief, culture, language and birth. Section 9(2) permits measures to be taken to protect or advance persons disadvantaged by unfair discrimination. This comes together with the duty of the state to 'respect, protect, promote and fulfil the rights in the Bill of Rights'. 43 This is complemented by the important role played by international human rights law in the development of the South African jurisprudence. Under s 39(1)(b) of the Constitution, courts must consider international law when interpreting the Bill of Rights. The Constitutional Court has frequently referred to international human rights conventions in pursuance of this duty. Most recently in Thubakgale, Majiedt J referred to the General Comment of the CESCR to elaborate the meaning of the right to housing in the SA Constitution. Similarly, in the School Nutrition case, the High Court referred to General Comment 15 of the Committee on the Rights of the Child (CRC), which sets out the crucial role of school feeding in the state's duty in relation to the rights to health and education in the CRC. 44 It also referred to CRC General Comment 19 on non-retrogression to underline the duty not to take deliberate retrogressive measures in relation to economic, social and cultural rights unless all other options have been considered.
There are several ways in which international human rights law is relevant. One is through the right to education, which is found in the ICESCR, CRC and CRPD, as well as the regional African Charter on the Rights and Welfare of the Child (ACRWC). While none of them expressly refer to ECCE, they all formulate the purpose of the right to education in terms that include the full development of the child's personality. There is now powerful evidence of the key role played by ECCE in the child's development, 45 strongly suggesting that ECCE is an implicit part of the right to education in all these conventions. 46 A second route is through the right to development. Both art 6(2) of the CRC and art 5 of ACRWC provide that States Parties must ensure the survival and development of the child to the maximum extent possible. 47 The CRC has stated that the right to survival and development can only be implemented in a holistic manner pulling together the rights to health and security, education and play. 48 Under the CRC, therefore, ECCE can be seen as one of the specific and potentially more directly implementable rights furthering the object of the duty to ensure the right to survival and development of the child. 49 This article focuses on the third route: the right to equality and non-discrimination in the CRC, ICESCR, CRPD and ACRWC. 50 It is further supported by the Convention on the Elimination of All Forms of Discrimination against Women (CEDAW), which is the only convention to expressly mention pre-school education. In this convention, States are required to ensure equality between men and women in, inter alia, pre-school education. 51 In the practice of the UN committees monitoring these conventions, as reflected in their Concluding Observations (CO), the right to equality is central to formulating the right to ECCE. This is because, as in South Africa, ECCE, where available, is limited to more privileged young children.
The following section sets out a framework of substantive equality, based on four dimensions, to assess the possibilities in domestic and international law in relation to ECCE. The article uses this framework to consider five questions. Firstly, given that poverty is the main factor driving inequality, can poverty be considered a ground of discrimination? Secondly, even if poverty is not a ground in itself, can race and genderand their intersection with povertybe utilised? Thirdly, what is the role of indirect discrimination? Fourthly, how should courts address resource-based justifications for limiting equality? And, fifthly, what role do and should positive duties play in promoting equality?

Substantive equality
The Constitutional Court has, from the start of the democratic era, expressed its commitment to substantive equality. 52 Substantive equality moves away from formal equality, which merely aims to treat likes alike. Instead, it accounts for the implications of historical, legal and social contexts on individuals and groups with certain protected characteristics, for example, by aiming for equality of opportunity (such as eliminating barriers to some groups) or equality of outcomes (such as requiring social benefits to be distributed fairly). Substantive equality has been elaborated in several ways. 53 This article uses a conception based on four mutually supportive dimensions in relation to outgroups: (i) redressing disadvantage; (ii) addressing stigma, stereotyping, prejudice and violence; (iii) facilitating voice and participation; and (iv) accommodating difference and achieving structural change. These need to be achieved simultaneously, as far as possible. 54 This framework has now substantially been accepted by the Committee on the Rights of Persons with Disabilities (CRPD) 55 in its interpretation of equality in the CRPD, as well as by the Supreme Court of India, 56 and the UK Equality Act 2010. 57 Below, we briefly elaborate on these four dimensions and their interaction.
Redressing disadvantage, the first dimension, is based on a recognition that the problem does not lie so much with the classification but the attached disadvantage. The right to substantive equality is not an egalitarian ideal that is fulfilled even if everyone is treated equally badly. Instead, it focuses on levelling up or extending benefits accorded to advantaged groups to outgroups. But disadvantage is not sufficient. It needs to be recognised that it is often stigma and stereotyping that fuel disadvantage. This is the second dimension, which reflects the close attention paid to dignity in the Constitution. To avoid the vagueness that potentially attaches to dignity, however, this dimension specifies the harms of stereotyping, prejudice, stigma and violence of all kinds. This extends disadvantage beyond acts against other individuals, which are in any case prohibited by the criminal law, such violence is fuelled and perpetuated by deep-seated stereotypes and stigmatic 'othering'. 58 The third dimension requires participation by those who are excluded and marginalised. 59 This is endorsed by Nancy Fraser, who sees parity of participation as the normative core of her conception of justice. 60 When past discrimination has blocked the avenues for political and social participation, equality laws compensate for this absence of political voice and open up channels for greater participation in the future. 61 This is particularly true for children. The first dimension (redressing disadvantage) combined with the participative dimension requires close attention to be paid to intersectionality. Kimberl e Crenshaw introduced intersectionality in 1989 to address the marginalisation of Black women in anti-discrimination law, as well as in feminist and antiracist politics and theory. Crenshaw showed how within these discourses, the prototypical representatives were white women and African American men. Since then, intersectionality has been applied across disciplines and globally. It is used here in its structural sense, referring to the synergistic disadvantage experienced by those at the intersection of multiple axes of power disadvantage. 62  sufficient simply to amplify the voice of elite members of a group. Those who are particularly disadvantaged and marginalisedparticularly at multiple intersectionsand who are able to speak to the needs of others in their position must be emphasised.
The fourth dimension recognises that inequality is more than the sum of many individual acts of prejudice but inheres in the structures of society, fuelled by imbalances of power. 63 To achieve the right to substantive equality therefore requires accommodation of difference and structural change. Behind all of these dimensions is the overriding need to redress inequalities in power. The focus should be on domination, or structures excluding or preventing people from determining their actions. 64 It is for this reason that the need to redress disadvantage both materially and of power, must interact closely with the need to enhance voice and participation and to bring about structural change.
In the light of this understanding of substantive equality, how could s 9 be utilised to reduce inequalities in relation to ECCE in South Africa, and how can this be buttressed by international human rights law? Given the absence of jurisprudence directly on this point, the following sections sketch potential pathways, rather than setting out an authoritative view.

Poverty as a ground of discrimination
It is clear that the strongest factor indicating lack of access to quality ECCE is poverty, which is not an express ground listed in s 9(3). Nevertheless, poverty exhibits many of the elements captured by the four-dimensional understanding of equality. Poverty in this context is more than income poverty. As Amartya Sen puts it: 'Human lives are battered and diminished in all kinds of different ways'. 65 Therefore a poverty index has been developed that includes poor health, lack of education, inadequate living standards, disempowerment, poor quality of work, the threat of violence, and living in areas that are environmentally hazardous. 66 All these elements are readily visible through the four-dimensional framework. As well as socio-economic disadvantage (the first dimension), the stigma often associated with poverty is in itself a harm. Starvation and deprivation, as well as creating material injury, can lead to deeply felt harms to dignity, both for those directly affected and for the society. As the judge put it in the School Nutrition case, 'A more undignified scenario than starvation of a child is unimaginable'. 67 Ruth Lister demonstrates that the ways in which the non-poor construct the poor as 'other' has a profound impact on how poverty is experienced. 68 The line between 'us' and 'them' is 'imbued with negative value judgments that construct "the poor" variously as a source of moral contamination, a threat, an 'undeserving' economic burden, or an object of pity'. 69 This was highlighted in Khosa v Minister of Social Development; Mahlaule v Minister of Social Development, 70 which held that permanent residents should be eligible for child benefits and pensions available to South African citizens. Mokgoro J emphasised that the consequences of exclusion were not only socio-economic but had a strong stigmatising effect, creating the impression that they were inferior to citizens and less worthy of social assistance. 71 The threat of violence is similarly recognised as an element of living in poverty.
The third dimension, lack of voice, is also apposite. People living in poverty are more likely to be excluded from participation in decisions that affect them, contrary to the third dimension. Most saliently, the voices of children, whose future participation in society might otherwise be blighted, must be amplified along with those of their families. Importantly too, poverty is systemic and only partially attributable to specific decisions by the state or private persons. For example, as discussed above, structural issues are central contributors to inequalities in ECCE. This is the fourth, transformative dimension.
Section 9(3) does not expressly include poverty as an enumerated ground. The grounds set out are, however, not exhaustive, and can be expanded by a court, although unenumerated ones do not benefit from the presumption of unfairness. 72 The Promotion of Equality and Prevention of Unfair Discrimination Act 4 of 2000 (the Equality Act or Pepuda) also does not refer to poverty within its long and inclusive list of protected grounds. Instead, the Act states that 'socio-economic status' is a 'directive principle' to which the Minister should give special consideration for inclusion. 73 Notably 'socio-economic status' is defined as 'a social or economic condition or perceived condition of a person who is disadvantaged by poverty, low employment status or lack of or low-level educational qualifications'. 74 However, thus far, socioeconomic status has not been given legislative force.
This leaves open the possibility of judicial extension. Under the Equality Act, prohibited grounds include the listed grounds, as well as any other ground where discrimination '(i) causes or perpetuates systemic discrimination, (ii) undermines human dignity; or (iii) adversely affects the equal enjoyment of a person's rights and freedoms in a serious manner that is comparable to discrimination on a ground in [the list]'. 75 In Social Justice Coalition v Minister of Police, a ground-breaking case in 2018, the Equality Court held that poverty was indeed a ground of discrimination under the Equality Act, finding that the state had discriminated on grounds of both race and poverty in the allocation of police human resources in the Western Cape. 76 In a statement with strong resonances for the provision of ECCE, Dolamo MJ stated: The dawn of democracy has not changed the lot of the people of Khayelitsha. They continue to live in informal settlements where the provisions of services are non-existent or at a minimum. This is more glaring where a comparison is made with the more affluent areas, mainly occupied by the privileged minority. 77 It may be arguable that the case rested on the intersection of race and poverty, and it remains to be seen how a case based purely on grounds of poverty would fare. However, the persistence of spatial apartheid flagged in that case, and by the Constitutional Court in Thubakgale v Ekurhuleni Metropolitan Municipality, 78 affects the provision of ECCE services in similar ways. Poverty and race are, unfortunately, highly likely to go together.
The move towards recognising poverty as a ground for discrimination is reinforced in international human rights law. As in South Africa, poverty is not included expressly in the long list of grounds found in the CRC and ICESCR. Nevertheless, in its 2009 General Comment 20 on equality and non-discrimination, the CESCR recognises that living in poverty may result in 'pervasive discrimination, stigmatisation and negative stereotyping which can lead to the refusal of or unequal access to the same quality of education and health care as others'. 79 It therefore states that economic and social situation should be regarded as included in the ground of 'other status' for vulnerable and marginalised social groups. This is carried through to the practice of the CESCR, the CRC and the CRPD, as revealed by our detailed examination of all the references to ECCE in the concluding observations (COs) these bodies issued between 2015 and 2020. 80 The committees express their concern regarding the cyclical relationship between education and poverty and point to the need to break this cycle by ensuring ECCE for poorer families. 81 Both the CRC and CESCR frequently note that poor child development is exacerbated by poverty, yet it is these children who are least likely to access ECCE. For example, the COs for the Dominican Republic (2015) and Suriname (2016) emphasise the importance of access to ECCE for children living in poverty, as well as single-headed households headed by adolescent girls. 82 This is true for both the Global North and the Global South. 83 A key issue is the intersectional relationship between poverty and those living in rural areas. Drawing on the importance of using the right to equality to extend existing provision to marginalised groups, the committees regularly express their concern that those in poorer settings are likely to have less access to ECCE. This can be seen in relation to Gambia, Lesotho, Ethiopia, 84 Ghana, 85 Angola and Guatemala. The committees also point out that good childcare can alleviate poverty 86 both in the short term by freeing up parental time for work and in the long term via better developmental and educational outcomes. 87

Gender, race and other grounds of discrimination
Even if poverty is not itself a ground of discrimination, the inequalities in relation to ECCE could be regarded as discrimination on grounds of race or gender. In the recent housing case of Thubakgale, 88 Majiedt J (dissenting) underlined the effect of continued spatial apartheid: 'Apartheid's spatial structures persist, and today continue to maintain race-and class-based inequities in access to resources and services across Johannesburg and surrounding areas'. 89 For him, the primary purpose of socio-economic rights is to 'promote substantive equality and human dignity, and also to undo the racialised system of poverty inherited from apartheid'. 90 There is no doubt that the lack of provision of quality ECCE in poorer areas maps onto ongoing racial segregation in South Africa. 91 There is also a significant gender component. Women and girls are discriminated against at a variety of levels that influence ECD, 92 demonstrating the need for a response along all four dimensions of the right to substantive equality. The interaction between stigma, stereotyping, prejudice and violence (the second dimension) and disadvantage (the first dimension), leading to ongoing structural inequality and intergenerational transmission of poverty (the fourth dimension), is particularly striking. There are multiple complex gender-related factors that impact ECCE. Although data does not suggest that fewer girls than boys attend ECCE in South Africa, we have seen above that boys are more likely to attend Grade R in a primary school setting than girls, suggesting that girls may be kept home while boys are sent to school at this early age. 93 These disadvantages are magnified by gender-based violence. In its 2016 CO on South Africa, the CRC expresses its deep concern at the high prevalence of gender-based violence against children, at home and in schools, especially in rural areas and informal settlements. These factors catalyse and magnify for pregnant learners, who are frequently expelled or suspended from school. In the 2013 Welkom case, 95 the Constitutional Court declared that school policies requiring pregnant learners to leave school would in principle amount to a breach of the right to gender equality. However, it refused to strike down the policies at issue in the case itself. Instead, the Court simply ordered the school governing body to review its pregnancy policies in light of the judgment. It also did not make any findings as to what provision the school should make for the pregnant learner, or for after the birth for the learner and her newborn. This has clearly not been sufficient. Despite the Welkom decision, both the CRC in 2016 and the CESCR in 2018 express concern in this regard. Notably, recommendations by both committees did not simply require adoption of policies. Recognising the structural nature of these inequalities, they also stated that positive measures should be put in place to ensure that pregnant teenagers and adolescent mothers receive support to continue their education. Thus, the CRC in 2016 recommended that South Africa 'ensure that pregnant teenagers and adolescent mothers are supported and assisted in continuing their education'; while the CESCR in its 2018 CO on South Africa states that the state should 'provide the necessary support services for pregnant adolescent girls, including measures to enable them to continue their education'.
Nevertheless, research suggests that despite Welkom and the CRC and CESCR recommendations, teachers perceive learner pregnancy negatively and do not believe they have the skills to effectively deal with pregnant learners. 96 Meanwhile, pregnancy of learners continues to rise. Recent DBE figures show that nearly 136,400 babies were delivered to women aged ten to nineteen in 2020, an increase of nearly 6,400 compared to the previous year. 97 The DBE's 2021 policy on the prevention and management of learner pregnancy in schools 98 provides helpful guidance on reducing learner pregnancy in the basic education sector, reducing stigma and discrimination and providing support for retention and re-enrolment. Strikingly absent, however, is holistic thinking on the simultaneous educational needs of both the learner and her child. This missed opportunity points to the need for robust research on the developmental outcomes of children born to very young mothers in South Africa.
Protection of pregnant learners is a consistent theme of the CRC, CESCR and CEDAW, which repeatedly couple the recommendation that pregnant learners be permitted to stay in school with the requirement that they be reintegrated after becoming mothers. 99 For example, in relation to Tanzania (2015) the CRC states: The Committee recommends that the State party take immediate measures to ensure the continued enrolment of adolescent mothers and girls who become pregnant during their school years and cease mandatory pregnancy testing. It also recommends that those girls who leave school because of pregnancy be supported and assisted in their re-enrolment and in the continuation of their education in mainstream schools. 100 Both the CESCR and CRC call on states to provide the services necessary to enable pregnant adolescents to continue their education. 101 They recognise that this is at least partly dependent on the support they are given to look after their children. The committees urge that adolescent mothers be given parenting support and childcare facilities. 102 For example in the 2020 Rwanda CO, the CRC urges the state to 'Ensure that pregnant girls remain in schools and reintegrate [them] into the schools [ … ] by providing counselling in parenting skills and childcare facilities'. 103 In an important further development, the West African Economic Community (ECOWAS) Court of Justice held that Sierra Leone's exclusion of pregnant learners from mainstream schools constitutes a violation of the right to education under the African Charter and the CRC. 104 There are further, complex interactions between the lack of ECCE opportunities and gender inequalities. Women disproportionately bear the responsibility for childcare during the closure of schools, ECCE programmes and childcare facilities due to Covid-19. 105 This is bound to affect their undertaking paid work and the hours they are able to work. Moreover, the overwhelming majority of ECCE workers in the poorest areas are Black women. 106 Much of the ECCE sector in these areas grew out of subsistence efforts of Black women who were restricted under apartheid from owning businesses: The highly informal nature of ECD sector provision provided an in-road for women into the labour market. But it is the most informal of businesses, and the poorest that have been hardest hit by the lockdown while the gender gap in earnings in the labour market has been further exacerbated. 107 Even before the pandemic, as the 2014 ECCE sector audit shows, average salaries were below the minimum wage, ranging from R1400 to R2000. 108 A 2020 report on the ECD sector during the pandemic found that 83 per cent of centres were not able to pay the full salaries of their staff, 109 91 per cent of which are women. 110 This is not just about the disadvantage of women or girls. Gender discrimination experienced by women deeply affects the children for whom they care. Women's lack of education, economic independence and childcare support either from fathers 111 or society 112 have a negative impact on parenting ability and child development, a problem the CESCR regularly emphasises. 113 If a mother has completed secondary education, her children are nearly five times more likely on average to attend an ECCE programme than children whose mothers have primary education or lower. 114 Moreover, maternal education is a key factor in providing informal early childhood education and preventing the intergenerational transmission of poverty. 115 Indeed, maternal education is one of the most reliable predictors of child development outcomes globally. 116 There are several other grounds of discrimination, where a similar line of reasoning could be used to that set out above. One is disability. 117 It is beyond the scope of this article to follow through this reasoning in detail, but it is clear that unequal access to ECCE for children with disabilities breaches all four dimensions of substantive equality. The Committee on the Rights of Persons with Disabilities has repeatedly recommended that African states should establish an inclusive education system at all levels, including pre-school, without discrimination. 118 It has also recommended that states should build and upgrade education facilities that are disability-sensitive and safe. 119 Similarly, the CRC has specified that states should ensure that children with disabilities have access to inclusive ECCE. 120

Indirect discrimination
As well as direct discrimination claims, there are some clear patterns of inequality that could be the basis of an indirect discrimination claim on grounds of poverty, and its intersection with race and gender. Although both the Constitution and the Equality Act prohibit indirect discrimination, 121 the term is not specifically defined. 122 However, it is well established in South Africa that 'a seemingly benign or neutral distinction that nevertheless has a disproportionate impact on certain groups amounts to indirect discrimination'. 123 There is no need for intention. How then could an indirect discrimination claim potentially be constructed? It is worth setting out how such a claim might be framed.
One source of indirect discrimination concerns lack of infrastructure. Even before the pandemic, the ECCE sector was characterised by infrastructural backlogs. As many as 45 per cent of centres do not have access to water inside the building, 40 per cent are not connected to a sewage system, and 33 per cent still make use of pit latrines. 124 It is difficult to pinpoint a particular law or policy that intentionally discriminates on grounds of poverty or race. Nevertheless, it might be possible to establish that ECCE policies that are apparently neutral in relation to these grounds have a disproportionate effect on those living in poverty, thereby potentially constituting indirect discrimination. One possible example might be the policy that in order to receive subsidies, an ECCE provider must be registered with the relevant department. In 2021, an estimated 60 per cent of all centresserving more than a million childrenwere not registered with the DSD. 125 It is highly likely that these children are poor and Black. 126 It is challenging for ECCE operators to become registered and therefore access subsidies. 127 An indirect discrimination claim would require government to look closely at the criteria for registration and ask whether there are other alternatives that can maintain appropriate standards but do not exclude centres in the poorest areas and serving the poorest children. As a 2020 report states: 'the question we need to ask ourselves is how government can design a registration system of which the majority are unable to form part'. 128 Similarly, centres are only able to claim subsidies for children with birth certificates, incentivising them to require a birth certificate for enrolment in the centre. While South Africa has made impressive progress in increasing rates of birth registration, 129 the poorest and most marginalised children still face obstacles in acquiring relevant documentation. Birth certificates are free in principle, but a family might need to make several trips to the relevant government office to ensure they have the right documentation. Given the cost of travel, the poorest families might be unable to make this journey. 130 This was of particular concern to the CRC in its CO for South Africa, where it points out that 'administrative and practical obstacles in obtaining birth registration, including punitive measures for late registration under the Births and Deaths Registration Act (Act No. 51 of 1992), may have negative and discriminatory impacts'. 131 This is especially problematic as 'possession of one's birth certificate is a rigid requirement for accessing social and child protection services'. 132 This falls especially heavily on children of migrant parents: because of the disproportionately strict conditions' for granting nationality, there are many children who are undocumented or whose births have not yet been registered. 133 Children of undocumented migrants are effectively unable to apply for a birth certificate given that parents are required to present their passports. 134 This means that ECCE providers serving migrant children are unlikely to be registered and therefore attract subsidies. 135 In the ground-breaking 2019 Phakamisa judgment, the Makhanda High Court held that the DBE's Admission Policy for Ordinary Public Schools not only breached the right to basic education but also the right to equality under s 9(3). 136 Notably, the Court found discrimination on the ground of 'documentary status', which, it held, was analogous to the enumerated grounds in s 9(3) because of the adverse effect on the learners' dignity. Although the Court relied on direct discrimination on the grounds of documentary status, which is arguably difficult to define, 137 it also emphasised the 'undeniable fact that the children affected [ … ] emanate from the vulnerable, poor black community'. 138 This means that the claim could also be framed in terms of indirect discrimination, allowing the principle to extend beyond the specific category of undocumented children. Such a framework would recognise the disproportionate effect on poor, migrant and rural children and other marginalised groups of requiring a birth certificate as a condition for access to education, and ask the government to establish an alternative, less discriminatory, means of managing admissions and resource distribution within the education system. This legal victory is significant in supporting undocumented learners to assert their right to basic education in South Africa. However, civil society organisations have submitted a report to CESCR in response to South Africa's country report on the Committee's COs, in which they demonstrate that the application of the judgment has been inadequate to enable undocumented migrant, refugee and asylum-seeking children to access basic education. 139 They point out that merely purporting to remove formal barriers is inadequate given the scale of the problem and the continued practices of excluding learners owing to lack of documentation. 140 Moreover, the Phakamisa judgment covers the right to basic education as provided by DBE, which has yet to explicitly include ECCE. With the recent move of function of ECD to DBE, further policy changes and alignments are expected. An extension of the Phakamisa judgment to cover undocumented learners in ECCE settings may not be automatic if ECCE is not recognised as part of a child's right to basic education.
A second potential source of indirect discrimination concerns the payment of fees. Fees are charged for more than 80 per cent of children aged zero to six years who are attending ECCE programmes not based at schools. 141 This contrasts with children in schools, where three-quarters do not pay any fees due to subsidies. The National Integrated Early Childhood Development Policy, published in 2015, developed a grant based on two components, an infrastructure component aimed at the development of low cost ECCE centres, and a subsidy component aimed at the nutritional and stimulation needs of children. 142 Only 33 per cent of centres currently receive the subsidy, with between 40 per cent and 44 per cent of centres receiving the subsidy in the three provinces with the highest child poverty rates. 143 These subsidies are crucial to the continued existence of many ECCE facilities in poorer areas, especially since so many families cannot afford the fees. This was starkly demonstrated when most of the subsidy was withdrawn during the Covid-19 lockdown. 144 The SA Childcare case itself helpfully held that the statutory duty to pay the subsidies had been breached and ordered the respondents to pay them immediately. 145 However, what it could not address was the low level of subsidies themselves, and their indirect discriminatory effect since the judgment was reserved for centres already registered. For 2022 to 2023, R1.1 billion was set aside for ECCE subsidies, of which R97 million was allocated for maintenance improvements and the construction of new low-cost centres. This represents an average growth of 1.7 percent, well below inflation. 146 Importantly, too, it is one of the lowest growth rates in the budget for basic education. From the R3.7 billion set aside for ECCE grants over the next three-year period, the government will be able to subsidise 600,000 more children. However, with an estimated 3.2 million children between zero and four years old still lacking access to ECCE, 147 this shortfall will only continue to increase. Challenging this lopsided allocation as indirectly discriminatory would be a bold strategy, but with the right facts, might be successful. For example, it might be argued that the decision to fund ECCE on a lesser scale than basic education, while apparently neutral on its face, has a disproportionate effect on poorer (predominantly Black) children who are 140 Ibid para 36. 141 Wills et al (note 2 above) 5. 142 Note 14 above. 143 DBE (note 20 above). 144  deprived of the important developmental benefits of ECCE, while children of wealthier families can avoid the detriment by attending fee-paying ECCE services. 148 Such an approach is further supported by international human rights law. COs of the monitoring committees consistently identify direct and indirect costs of ECCE as barriers deterring poorer and marginalised communities from accessing ECCE. Indirect costs include transport costs, especially in poorer and rural communities, teaching materials 149 and birth registration. 150 This is further borne out by the data. A UNESCO study shows that, across countries, the burden on household spending is much greater in relation to pre-primary education than other levels of education. The report cites studies from various parts of Africa and South Asia that show that while poor families face a very high financial burden in the provision of pre-school education, they are also at risk of low-quality service from private schools. 151

Justification: Adjudication of resources
Even if a prima facie case of indirect discrimination can be established, there is still an opportunity for the respondent to demonstrate that it is justifiable, in that the government is using proportionate means to achieve a legitimate end. In principle, this should be an exacting standard, given the very high cost both to young children and society as a whole of continuing inequality in ECCE provision. Florencia Lopez Boo, Jere Behrman and Claudia Vazquez simulated the present discounted value of losses in future income of pre-primary closures due to Covid-19 lockdowns in over 140 countries with the combined populations of 6.4 billion people and 323 million pre-school-age children. They found that millions of children of pre-primary school age are likely to suffer considerable earnings losses over their lifetimes due to preschool-programme closures. This is before considering other effects, including a longterm decline in quality and availability of pre-school programmes, and the effects on primary caregivers. 152 Courts have traditionally been cautious about adjudicating on issues with resource implications, on the assumption that such decisions should be made by legislators, who are accountable to the electorate, and the executive, who are more competent in dealing with polycentric decisions. Indeed, whereas courts in many jurisdictions have felt comfortable with adjudicating equality claims based on status or protected characteristics, they have generally eschewed a role in relation to economic inequalities. 153 However, as we have seen, status-based inequalities, such as gender or race, are inextricably linked with economic inequalities, and attempts to confine judicial intervention to the former have inevitably been ineffective. This is particularly glaring for inequalities in ECCE provision, when low quality or no ECCE services cement and exacerbate inequalities in basic education and then in later life.
Catherine Albertyn argues that s 9(3) could be used to expand the scope of existing opportunities, or alter their distribution but, arguably, not authorise new social benefits. 154 Even if courts are deferent on the budgetary allocation, there is scope to examine existing budgets to determine whether they are being used as effectively as they should and with the right priorities. In a series of cases brought by the Legal Resources Centre to establish that failure to provide proper schools to replace the mud structures erected by local communities was a breach of the right to education, one of the key issues was to achieve accountability for budgets already allocated for this purpose. 155 There may also be scope to require detailed justification for a budget that indirectly discriminates on grounds of poverty, race and, potentially, social origin. In the context of ECCE, there is already a well-developed policy: courts in this context would not need to develop a policy but rather require that the indirectly discriminatory effect of insufficient resource allocation be justified on a stringent proportionality test. The need for courts to insist on full accountability for their budgeting preferences was endorsed by the Constitutional Court in Blue Moonlight, when it held that it was not 'good enough for the City to state that it has not budgeted for something, if it should indeed have planned and budgeted for it in the fulfilment of its obligations'. 156 Given that the right to equality under s 9 is not subject to available resources, as is the case for the right to housing which was at issue in Blue Moonlight, there should be an even stronger requirement for accountability in this context. 157 This was further endorsed by the CRC in its CO for South Africa, when it noted, with concern, continued wide disparities according to economic status, race and geography in access to quality education alongside 'the persistence of the uneven distribution of public resources, resources being allocated to address less critical issues rather than the most urgent ones, and the lack of transparency in the management of funding in the education system'. 158 This prompted the Committee to recommend that South Africa improve the transparency, efficiency and accountability of the management of the education budget. Importantly, too, consistent with the third dimension of substantive equality (voice and participation), it recommended 'the active and meaningful participation by children and civil society organisations in the development of the budget and through monitoring and evaluation of its implementation'. 159 This is echoed for other countries, where the Committees consistently require resource allocation to achieve greater equality. For example, in relation to Ghana, the CRC emphasised the need to allocate 'sufficient financial resources for the development and expansion of early childhood education in rural areas, and continue strengthening the efforts to efficiently implement the Early Childhood Care and Development Policy (2004)'. 160  3.6 Positive duties to promote equality A further route towards using the right to equality to reduce the inequalities so evident in the provision of ECCE is through positive duties to promote equality. Such duties are proactive, requiring the state take the initiative in promoting equality, rather than waiting for individuals to bring claims of discrimination. Proactive duties require the structural causes of inequality to be addressed, recognising that that inequality extends well beyond individual acts. 162 In this way, proactive duties are a core part of achieving all four dimensions of substantive equality, moving beyond the individualistic focus of anti-discrimination law, to require structural changes that redress disadvantage and address stigma, stereotyping, prejudice and violence. It is crucial, too, that proactive duties ensure those affected participate in decisions, and that their voices are heard. It should be stressed that proactive measures need not entail 'affirmative action' or express preference policies. Rather, they require the state to 'mainstream' the right to equality, for example, through equality impact assessments to check all legislation and policies for their impact on inequality and actively devise measures which promote equality. 163 Under s 7(2) of the Constitution, the state must respect, protect, promote and fulfil the rights in the Bill of Rights, entailing that there is a positive duty to fulfil the right to equality in s 9. This is augmented by s 9. Section 9(1) provides not only that everyone should have 'equal protection' of the law, but also 'equal benefit' of the law. This suggests that the negative duty of protection against infringements of the right should be augmented by the positive duty to ensure equal benefit of the law. Furthermore, s 9(2) states that, to promote the achievement of equality, the state may take measures designed to 'protect or advance' persons disadvantaged by unfair discrimination. This provision has generally been regarded as permitting the state to implement affirmative action provisions, such as those under the Employment Equity Act 55 of 1998, which require designated employers to ensure that suitably qualified Black people, women and disabled people are equitably represented in the workforce. Although this creates preferences on grounds of race, gender and disability, it does not breach the prohibition of unfair discrimination on these grounds, as long as it is designed to protect or advance these groups. Section 9(2) on its face is permissive rather than mandatory. Nevertheless, Albertyn has persuasively argued that recent case law suggests that 's9(2) might be more than a defence for the state's positive measures, and can foster mandatory positive action'. 164 In particular, 'a failure to develop and implement equality-related constitutional and legislative objectives might lead courts to impose positive duties on government to act in a manner to achieve them'. 165 As mentioned above, such positive duties need not constitute a putative breach of the prohibition on unfair discrimination in s 9(3), and this is particularly true in relation to ECCE. This means that the mandatory nature of the duty can be found as much in s 9(1) and s 7(2) as in s 9(2).
The importance of proactive duties was recognised in the Equality Act, which gives the promotion of equality as one of its explicit objectives. 166 It is, however, unfortunate that Chapter 5 of the Act, which elaborates on the duty of the state and of all persons to 'promote and achieve equality', 167 has not been brought into force. Chapter 5 has much potential. The positive duty in this section includes a duty on Ministers and public contractors to prepare and implement equality plans. 168 The Act defines equality as including de facto equality and equality of outcomes. 169 Given the commitment in the Constitution to substantive equality, the promotion of equality in this context should include redressing disadvantage, addressing stigma, stereotyping and violence, facilitating participation, and achieving structural change. 170 The promotion and achievement of equality under the Act should therefore include proper budgetary allocations to redress the disadvantage of young children, with positive spin-offs for furthering race and gender equality, and reduce the stigma and prejudice accompanying such disadvantage. It should make sure to consult with and listen to those who are already involved in providing ECCE to ensure effective change. The ultimate aim is to break the cycle of poverty into which ECCE absence locks children.
The Equality Act is currently going through a process of amendment, and it is hoped that Chapter 5, possibly in a strengthened form, would finally become law. Under the subsidiarity principle developed by the SACC, 'where legislation has been enacted to give effect to a right, a litigant should rely on that legislation in order to give effect to the right, or, alternatively challenge the legislation as being inconsistent with the Constitution'. 171 It may be worth waiting for the process of amendment to be completed. In any event, any challenge to the Equality Act in this respect would need to be a full facial challenge. 172 164 Albertyn (note 154 above) 229. 165 Ibid. 166 Pepuda s 2(b)(ii). 167 Ibid s 24. 168 Ibid s 25 -27. 169 Ibid s 1(ix). 170 See s 149 of the UK Equality Act, which includes these dimensions in the aims of the positive duty and s 158 and s 159, which set out these as criteria for legitimate positive action. 171  Positive measures are an integral part of the work of the monitoring committees, which regularly make recommendations as to the actions States Parties should take to achieve equality in relation to ECCE. For example, in its CO for Estonia, the CRC recommended that the State Party 'Strengthen the system of quality pre-school education and increase availability and affordability of pre-school education, including for children from marginalised families'. 173 In relation to Guatemala, the CRC recommended that the State Party '(a) Adopt a strategy aimed at increasing the enrolment of children in primary, secondary school and pre-school, in particular those living in areas with high levels of multidimensional poverty'. 174 Inclusion of those affected and listening to their voices is also stressed. For example, in its COs for Slovakia, the CESCR recommended that the State should: Consider, in consultation with the relevant stakeholders, including families from across different geographical areas and ethnic and socioeconomic groups, how to effectively balance provisions for parental leave, investment in preschool education and support for families to balance family and working responsibilities. 175 The CRC was even more specific in relation to Botswana, recommending that the State 'reintegrate pregnant girls and adolescent mothers into the mainstream school system by providing counselling in parenting skills and childcare facilities'. 176 In Kenya the CRC stated 'there is still a large unmet need among parents, including teenage parents, for parenting skills and education' and that the state should provide 'further support to families, including teenage parents, in the form of family counselling, parenting education and the provision of financial allowances'. 177

Conclusion
Covid-19 has exacerbated vulnerabilities in ECCE provision in South Africa with nearly 3.2 million children under five without any access to a programme and many more with access to only poorly resourced, low-quality, improvised services. 178 As is often the case, it is the most vulnerable who face the greatest challenges to access and quality perpetuating their vulnerability. This article has explored how the right to equality in the Constitution can be used to address these gaps in provision in ECCE. We have argued in the context of the Constitution and international law that within a framework of substantive equality, poverty, gender and racial inequalities are potential grounds for discrimination both directly and indirectly in relation to ECCE. We further propose that resource-based justifications for limiting the right to equality are inadequate when budgets allow for resources to be unequally distributed contravening a government's positive duty to fulfil the right to equality.